Rol IA · Oferta abierta

LLM Fine-tuning Engineer — domain adaptation

Índice de mercado

Enfoque: Voice AI

Paris · 50-100 employees · Enterprise · GenAI: No

Tiempo completo
Senior
Model post-training
Remote
Remote · US-East compatible
Toronto, Canada
CAD 150k–190k

Este anuncio está pensado para talento nativo en IA: habilidades y herramientas claras para saber si encajas antes de aplicar — y para reducir descartes por desajuste.

3 habilidades2 herramientas

Aplicar ahora Buscar talento afín

Compartir este rol

De un vistazo

Cómo trabajarás, dónde y fechas clave del anuncio.

Empleo: Tiempo completo
Modalidad: Remote
Ubicación: Remote · US-East compatible
Compensación: CAD 150k–190k
Nivel de experiencia: Senior
Equipo: Model post-training
Beneficios

Sobre el rol

Qué busca el equipo y cómo encaja el puesto.

Snips is hiring a Senior LLM Fine-tuning Engineer — domain adaptation on the Model post-training team (Remote · Toronto, Canada). This is a hands-on role in LLM & generative AI—not a generic “AI enthusiast” post.

You will own LLM features, retrieval, agents, and evaluation end to end: problem framing, offline evaluation, safe rollout, and iteration from production signals. Success is measured with clear metrics (quality, latency, cost, or business KPIs)—not slide decks alone.

Remote role aligned to US-East hours but employed via our Canadian entity. You will specialise in domain adaptation, preference tuning, and refusal behaviour for regulated enterprise assistants.

Travel to customer workshops in major US cities 4–6 times per year.

What you will do

Design data mixes (instruction, preference, safety) with traceable provenance.
Run large fine-tunes and LoRA/QLoRA recipes with reproducible configs.
Partner with eval teams to gate promotions on domain-specific suites.
Optimise training infra cost (spot, multi-node, checkpointing).
Document playbooks for customers running private fine-tunes on our stack.

What we look for

5+ years ML with 2+ focused on LLM post-training.
PyTorch expert; experience with TRL/Axolotl-class tooling or internal equivalents.
Clear communication with policy teams on safety trade-offs.

Nice to have

Mixture-of-experts or long-context specialty.
Hardware procurement / cluster negotiation exposure.

Habilidades que buscamos

Qué implica cada capacidad en este rol — no solo una lista de palabras clave.

Python
Production Python for data pipelines, model code, APIs, and automated tests—not notebook-only workflows.
OpenAI API
Calling frontier APIs with retries, cost controls, prompt versioning, and safety filters.
PyTorch

Herramientas y stack

Plataformas y productos mencionados para este rol (APIs, IDEs, BBDD vectoriales, CRM, diseño, etc.).

AWSTool
Regular use of AWS in the LLM & generative AI stack—configuration, debugging, and team conventions.
DockerTool
Containerized services and reproducible dev environments for ML and LLM workloads.

Encajas muy bien si…

Adaptado a esta oferta — úsalo para comprobar tu encaje antes de aplicar.

You can show evidence in your application: Describe a catastrophic forgetting issue and your mitigation.
Several of these show up in recent shipped work—not only on your CV: Python, OpenAI API, PyTorch, AWS, and Docker.
You have owned LLM features, retrieval, agents, and evaluation in production: debugging live issues, running postmortems, and iterating from real user or business signals.
You communicate clearly in writing, document decisions, and can overlap with the team’s core hours for pairing and incident response.
You explain trade-offs (quality, cost, latency, safety) in plain language—with numbers or examples, not buzzwords.
You read the full listing (comp: CAD 150k–190k, benefits in At a glance) and your expectations align.

Aplicar

Solo cuentas candidato: añade tu titular y bio en , y opcionalmente un enlace o PDF de CV para que los reclutadores vean tu CV completo. Luego escribe un breve pitch — las candidaturas usan el correo de tu cuenta.

Lo que pide la empresa

Describe a catastrophic forgetting issue and your mitigation.

Inicia sesión con cuenta candidato para aplicar

Regístrate, añade titular y bio; luego podrás adjuntar enlace o PDF de CV en Mi cuenta antes de aplicar. Así las candidaturas quedan ligadas a tu perfil.

Iniciar sesión Crear cuenta candidato

Sugerencias

Talento IA que podría interesarte

Perfiles ordenados por solape con habilidades y herramientas de este rol — útil si contratas equipo o comparas candidatos.

Aún no hay coincidencias cercanas en el directorio — prueba con tus filtros.

Ofertas similares

Otros anuncios que comparten habilidades o herramientas — útil para stacks comparables o alternativas.

LLM Fine-tuning Engineer — domain adaptation

Talento IA que podría interesarte

Ofertas similares

Lead PyTorch Engineer — LLM fine-tuning & distributed training

Part-time ML Researcher — time-series + LLM hybrids

Senior LLM Engineer — enterprise RAG

Senior ML Engineer — payments fraud

Senior AI Developer — HR copilots